Generative AI image solutions, like Midjourney, are a type of artificial intelligence that can create entirely new images based on text descriptions, based on, e.g. the Diffusion Model. You can generate all kinds of still images, from basic lines to hyper-realistic images. It is based on your imagination.
Here's a breakdown of how they work:
Natural Language Text Input > AI Processing > Image Generation > Refinement (if needed):
Benefits of Generative AI Image Solutions:
Spark Creativity: These tools can help you visualise ideas; explore creative concepts that might be difficult to achieve with traditional methods; and augment your creative process.
Rapid Prototyping: You can quickly generate multiple image variations to test different ideas, experiment with diverse possibilities, and get a feel for what works best.
Non-Artistic Skills Accessibility: Even if you don't have artistic skills, you can create high-quality images using text descriptions.
Here are the examples of use cases:
Illustrators and designers can use these tools to generate initial concepts for characters, environments, or products
Create eye-catching visuals for social media posts, presentations, or marketing campaigns
Generate images to illustrate stories, poems, or other creative writing projects
Explore different design concepts and ideas for new product
Create unique and personalised artwork based on your imagination
Prompt 101
Midjourney: “A prompt is a short text phrase that Midjourney Bot interprets to produce an image”.
Prompt Design
Google Could stated that Prompt design is “the process of creating prompts that elicit the desired response from language models. Writing well structured prompts can be an essential part of ensuring accurate, high quality responses from a language model.” Prompt Design is analogous to a cocktail process - a mix of ingredients.
PROMPT: Subject + Medium + Environment + Lighting + Colour + Mood + Composition
Here are the examples of this image:
Long Version Prompt: A hyper-realistic oil painting of an elegant Siamese cat in the garden, surrounded by lush greenery and blooming flowers, with intricate details on its fur and expressive eyes, illuminated by natural sunlight. The background features wooden fences and tall grasses, creating depth and texture to enhance the scene's realism.
Short Version Prompt: hyper-realistic oil painting of an elegant Siamese cat in the garden, greenery, flowers, with intricate, eyes, illuminated, sunlight, features wooden fences, grasses
Midjourney Prompt
There are three methods that you can prompt in Midjourney:
Method 1: Learn how to prompt
Learning to use the correct term? It is NLP - Natural Language Processing. It is learnt from the existing data. In the creative world, you need to learn terms, jargon, vocabulary, etc If you want to generate ‘stunning’ images you have to learn new vocabularies in each area,
Realistic Image VS Photorealistic Image
Be specific -
Hotel Owners VS Thai Hotel Owners
Generative AI Image is not conversational AI, it is image processing AI. Instructional AI.
Instead of prompting: can you help to generate an image of a beautiful lady?
Prompt: You have to prompt: a black lady, beautifully slim, blue and big eyes, wearing pencil skirts, portrait style
photorealistic image…
Be short and concise
Not necessarily to prompt a long instructional sentence, Try to compose not a sentence, but a short phrase in order
—- Subject, Medium, Detail, Context, Output style, Image Size —
Method 2: Use Generative AI Text to help
Use ChatGPT / Gemini helps you to generate a prompt by telling your descriptive story, and then bring the prompt to generate in Generative AI Image.
Method 3: Prompt Platform
You can learn from’ Prompt Platform’ - learn from others.
To learn more about Midjourney Prompt, please look at Midjourney Document: https://docs.midjourney.com
#AI #GenerativeAI #Midjourney #Creativity #Promptdesign
#TheCellUnit #TCURightBrainPartnership